Methodology

The short version

Most newsletters don't publish their subscriber counts or revenue. We estimate them from public signals — archive pages, engagement patterns, pricing pages, and platform metadata — and label every number with a confidence level. Verified numbers come from creators directly or from platform APIs. Everything else is an estimate, and we tell you which.

Confidence levels

Every subscriber count and revenue figure on the site carries one of four confidence labels. They map to concrete error ranges:

Ranges are typical, not guaranteed. We use them to communicate uncertainty, not as a contractual SLA. A Verified number is point-in-time accurate; a Low estimate may be off by a factor of two.

Subscriber estimation

Subscriber counts come from one of these sources, in order of preference:

1. Creator export. CSV uploaded by the verified owner. Verified.
2. Platform API. Ghost Admin API or Beehiiv API for connected accounts. Verified.
3. Public display. Counts a newsletter chooses to show (e.g. X subscribers on a Substack profile). High to Medium confidence depending on roundedness — Substack's inbox API returns rounded thousands for most newsletters, so we treat exact-looking counts higher than round-thousand counts.
4. Engagement estimation. When no count is published, we estimate from comment volume, like counts, and posting cadence relative to known peers in the same category. Low confidence (±50–75% depending on signal density).

Estimation accuracy varies by platform. Substack is best — the archive ladder gives us strong signal. Ghost and Beehiiv estimates currently run ±50–75% for newsletters without API connections; tightening that range is on our roadmap.

Paid vs free split

For paid newsletters, we estimate the share of subscribers on the paid tier. Where the platform reports a paid count directly (Ghost API, Beehiiv API, creator export), we use it as-is. Otherwise we apply category-aware paid conversion ratios to the total subscriber count.

Creators can override our paid-subscriber estimate from their settings page once verified. Manual overrides take precedence over estimates and are surfaced as Verified.

Subscription revenue

Subscription revenue is computed as:

paid_subscribers × effective_monthly_price × (1 − platform_cut)

Effective monthly price

We scrape every paid tier from the newsletter's pricing page and use the cheapest plan in each billing interval — not the founding-member or premium tier. When both monthly and annual plans exist, we blend them assuming 70% of paid subscribers choose annual — an industry estimate based on public commentary from Substack and Ghost rather than a directly measured figure. We use it as a default; per-newsletter actuals will vary.

Platform cuts

We deduct platform fees from creator-facing revenue. Substack takes 10%; Ghost and Beehiiv take 0% on subscription revenue (they charge platform fees instead, which we don't deduct because they're a fixed cost rather than a percentage of revenue).

Sanity caps

Three caps prevent supporter tiers, founding-member plans, and corrupt pricing data from inflating estimates:

• $100/mo cap on effective price. Filters Ghost $500/mo supporter tiers and similar.
• 1.5× annual/monthly ratio cap. If annual ÷ 12 exceeds monthly × 1.5, the annual is treated as a premium tier and discarded.
• $50/subscriber/month revenue cap. Final guard against course prices, enterprise tiers, or one-time donations stored as recurring subscription pricing.

Ad revenue

Newsletters with sponsorships earn ad revenue alongside (or instead of) subscriptions. We estimate it as:

subscribers × open_rate × sponsored_editions × CPM ÷ 1000

Defaults

• Open rate: 40%. Newsletter industry average across reporting platforms. We don't track per-newsletter open rates without a platform connection. Note: reported open rates include Apple Mail Privacy Protection inflation (Apple pre-fetches images, registering as opens). A more conservative 35% would reduce ad-revenue estimates by ~12% — buyers reading these numbers should treat them as the optimistic end of a 35-40% band.
• Sponsored editions: capped at 8/month, discounted by a 50% fill rate. A 4×/week newsletter publishes 16 editions, but realistically monetizes ~4 of them.
• CPM by category. Tech/Business/Finance/AI/Crypto/Marketing: $35. Politics/News: $25. Lifestyle/Culture/Entertainment/Design: $20. Default: $15. These are conservative floors; premium properties (named brands, established enterprise audiences, exclusive placements) earn meaningfully higher CPMs that we can't capture without negotiated-rate visibility. Treat the resulting ad-revenue numbers as a lower bound for top-tier properties.

Eligibility

We only estimate ad revenue for newsletters that meet at least one: known sponsorships in our index, 10K+ subscribers, or ad-native platform (Ghost or Beehiiv with no paid tier). This avoids inflating revenue for hobby newsletters that never run sponsorships.

What we don't measure

• LinkedIn revenue. LinkedIn newsletters report followers, not paid subscribers, and the platform doesn't support paid tiers or sponsorships at the publication level. We track audience size only.
• Off-platform revenue. Courses, books, consulting, and merch sold alongside a newsletter are not in our model. The number you see is the newsletter business in isolation.
• Negotiated deal prices. Our CPM defaults are category averages. Negotiated rates for premium properties (Lenny's, Stratechery, Morning Brew) are typically higher and we don't have visibility into them.
• Refunds, churn, and promotional pricing. Subscription revenue is steady-state at the listed price. Trials, intro discounts, and annual-payment lumpiness aren't modeled.

Verifying or correcting a number

Newsletter owners can claim their listing and correct any number we've estimated. Verification is via DNS TXT record, meta tag, or email code, after which manual overrides for subscriber/paid counts and a CSV export upload promote the listing's confidence to Verified.

If you're a newsletter owner and our estimate looks wrong: claim your listing or email support@newsletterinsights.io. We update Verified data within one business day.

If you're a researcher, advertiser, or investor and want to understand the methodology behind a specific number, we'll walk you through how it was derived. Same email.

Updates

We tighten estimation methods over time. When a methodology change materially affects historical numbers (e.g. a recalibration of Ghost/Beehiiv estimates), we'll note it here with a date and the affected newsletters.